Reliability of Glucometer Blood Glucose Readings in Neonates

Shayan Pavaskar, Siddhuraj Rangesh, Dhanam Venkatachalam Suresh, Santhanakrishnan Ramakrishnan

Abstract

Aim: To assess the agreement between bedside glucometer readings and laboratory blood glucose estimation in neonates

Materials and Methods: This prospective observational study was performed in the tertiary NICU of a private hospital in a tier 2 city in south India. The study duration was 5 months, and it included 100 neonates who required blood glucose testing and an intravenous cannula insertion as part of their routine clinical care. Venous blood samples of the neonates were sent to the laboratory for plasma glucose estimation and simultaneously tested for whole blood glucose using the Accu-Chek® Performa glucometer.

Results: There was a lack of agreement between the 2 methods, with a difference of up to 40 mg/dL. The venous blood glucose estimation using a glucometer may be 33 mg/dL above or 13 mg/dL below the values of plasma glucose concentration obtained from the laboratory test. The 95% confidence interval for the upper limits of degree of agreement was 28.7 to 36.6 mg/dL, and for the lower limits of agreement, it was 8.7 to 16.6 mg/dL.

Conclusions: Not all glucometers can be relied upon to accurately monitor neonatal blood glucose. Only glucometers approved for use in the neonatal population must be used for desired accuracy.